Blockchain technology is a sustainable technology that offers a high level of security for many industrial applications. Blockchain has numerous benefits, such as decentralisation, immutability and tamper-proofing. Blockchain is composed of two processes, namely, mining (the process of adding a new block or transaction to the global public ledger created by the previous block) and validation (the process of validating the new block added). Several consensus protocols have been introduced to validate blockchain transactions, Proof-of-Work (PoW) and Proof-of-Stake (PoS), which are crucial to cryptocurrencies, such as Bitcoin. However, these consensus protocols are vulnerable to double-spending attacks. Amongst these attacks, the 51% attack is the most prominent because it involves forking a blockchain to conduct double spending. Many attempts have been made to solve this issue, and examples include delayed proof-of-work (PoW) and several Byzantine fault tolerance mechanisms. These attempts, however, suffer from delay issues and unsorted block sequences. This study proposes a hybrid algorithm that combines PoS and PoW mechanisms to provide a fair mining reward to the miner/validator by conducting forking to combine PoW and PoS consensuses. As demonstrated by the experimental results, the proposed algorithm can reduce the possibility of intruders performing double mining because it requires achieving 100% dominance in the network, which is impossible.

High blood pressure (BP) may lead to further health complications if not monitored and controlled, especially for critically ill patients. Particularly, there are two types of blood pressure monitoring, invasive measurement, whereby a central line is inserted into the patient’s body, which is associated with infection risks. The second measurement is cuff-based that monitors BP by detecting the blood volume change at the skin surface using a pulse oximeter or wearable devices such as a smartwatch. This paper aims to estimate the blood pressure using machine learning from photoplethysmogram (PPG) signals, which is obtained from cuff-based monitoring. To avoid the issues associated with machine learning such as improperly choosing the classifiers and/or not selecting the best features, this paper utilized the tree-based pipeline optimization tool (TPOT) to automate the machine learning pipeline to select the best regression models for estimating both systolic BP (SBP) and diastolic BP (DBP) separately. As a pre-processing stage, notch filter, band-pass filter, and zero phase filtering were applied by TPOT to eliminate any potential noise inherent in the signal. Then, the automated feature selection was performed to select the best features to estimate the BP, including SBP and DBP features, which are extracted using random forest (RF) and k-nearest neighbors (KNN), respectively. To train and test the model, the PhysioNet global dataset was used, which contains 32.061 million samples for 1000 subjects. Finally, the proposed approach was evaluated and validated using the mean absolute error (MAE). The results obtained were 6.52 mmHg for SBS and 4.19 mmHg for DBP, which show the superiority of the proposed model over the related works.

Despite significant advances in natural language processing, conversational AI systems face persistent challenges in maintaining extensive and contextually coherent dialogues, particularly regarding long-term memory management. This literature review synthesizes current approaches to memory architectures in conversational AI, examining the transition from basic dialogue agents to more sophisticated, agentic frameworks.We analyze how vector databases and Retrieval-Augmented Generation (RAG) address fundamental challenges in storing and retrieving conversational context, maintaining system responsiveness, managing user-specific data ethically, and integrating domain-specific information. Through systematic review of papers, we identify critical limitations of vector embeddings in capturing extended conversational context, particularly in agentic domains requiring semantic, episodic, procedural, and emotional memory. We evaluate how RAG frameworks can augment vector databases to handle memory-intensive tasks requiring real-time updates and domain-specific knowledge integration. Furthermore, we examine alternative architectures including knowledge graphs, finite state machines, and hybrid solutions, highlighting the data quality and ethical challenges that must be addressed for scalable, reliable AI memory management. Our analysis provides a structured framework for understanding memory evolution in conversational AI, identifies gaps in current RAG solutions, proposes hybrid memory designs, and outlines future research directions emphasizing cross-domain applications in agriculture.

Efficient detection and rectification of metal components conditions during manufacturing and post-processing manufacturing are crucial for quality control in industries. This paper describes a lab-scale integrated system for real-time and auto-mated metal edge image detection using YOLOv5...

Agriculture requires accurate, location-specific information that would need the power of advanced Retrieval-Augmented Generation (RAG) models. To this end, we perform an experimental analysis on how integrating re-ranking strategies and in-memory computing into RAG models might affect performance on small agriculture question-answering (QA) datasets. This method envisages to enable real-time ground-truth kind of answers for agro-informatics sake, the proposed approach is to assist enhance document relevance and lower response latency. We trained the system on a large-scale agriculture QA dataset using high-level components like the Sentence Transformer for embedding generation, FAISS for fast vector search and a pre-trained language model for response generation. This is to keep the documents returned highly relevant, and zero-shot classification was used for re-ranking techniques. The efficacy of their algorithm across a range of QDMR transformation tasks was evaluated, and the experiment evaluation showed that rereading did not significantly increase performance over baselines. But the in-memory computing with FAISS greatly reduced retrieval latency which makes it appropriate for real-time applications in agriculture QA systems.